PI: Farrah Moazeni
University: Lehigh University
Industry partner: City of Bethlehem
Municipal water utilities are increasingly expected to operate complex distribution networks in real time using limited sensing, aging infrastructure, and incomplete system visibility. While digital twin technologies are often proposed as a solution, most existing approaches assume dense instrumentation and high-fidelity models that are not available in practice for many Pennsylvania municipalities. As a result, utilities lack reliable estimates of internal network states, such as unmeasured pressures, flows, and system anomalies, needed to support informed operational and investment decisions.
This project will develop and pilot an observability-driven digital twin of the City of Bethlehem’s water distribution network that is explicitly designed around the City’s existing sensors, data availability, and operational constraints. The project will first perform a systematic observability analysis of the water network to identify which states can be reliably inferred from current measurements and where estimation uncertainty is unavoidable. Building on this analysis, we will then design robust state estimation algorithms that combine physical network models with operational data to continuously estimate key system states in real time, even under sparse and noisy sensing conditions.
The resulting digital twin will be a living, continuously updated representation of the water network that provides utility operators with actionable situational awareness rather than static visualization. The twin will support early detection of abnormal system behavior, improved understanding of network performance, and data-driven guidance for future sensor placement and infrastructure upgrades. Importantly, the framework will be designed to integrate with existing utility workflows without requiring major hardware investments or cloud-based dependencies.
We will closely collaborate with the City of Bethlehem to validate the digital twin on a real municipal system and advance the technology from proof-of-concept to a deployable tool. The outcomes will directly benefit Pennsylvania by improving water system reliability, operational efficiency, and resilience, while establishing a transferable methodology that can be adopted by other municipalities across the Commonwealth.